stating that “[a] full trial de novo in any ERISA dispute where 13 there was a genuine dispute of fact as to whether the individual qualified for a benefit would 14 undermine” the policies underlying ERISA; “[t]he means that suggests itself for accomplishing 15 trial of disputed facts, while preserving the value of the fiduciary review procedure, keeping costs 16 and premiums down, and minimizing diversion of benefit money to litigation expense, is trial on 17 the administrative record, in cases where the trial court does not find it necessary under Mongeluzo 18 to consider additional evidence”
How later courts described this case
- stating that “[a] full trial de novo in any ERISA dispute where 13 there was a genuine dispute of fact as to whether the individual qualified for a benefit would 14 undermine” the policies underlying ERISA; “[t]he means that suggests itself for accomplishing 15 trial of disputed facts, while preserving the value of the fiduciary review procedure, keeping costs 16 and premiums down, and minimizing diversion of benefit money to litigation expense, is trial on 17 the administrative record, in cases where the trial court does not find it necessary under Mongeluzo 18 to consider additional evidence”
- explaining that in a trial on the administrative record, “[the] district judge will be 20 asking . . . as he reads the evidence, . . . whether [the plaintiff] is disabled within the terms of the 21 policy” and may “evaluate the persuasiveness of conflicting testimony and decide which is more 22 likely true”
- explaining that in a trial on the administrative record, “[the] district judge will be asking . . 19 . as he reads the evidence, . . . whether [the plaintiff] is disabled within the terms of the policy” 20 and may “evaluate the persuasiveness of conflicting testimony and decide which is more likely 21 true”
- explaining that in a trial on the administrative record, “[the] district judge will be asking . . 4 . as he reads the evidence, . . . whether [the plaintiff] is disabled within the terms of the policy” 5 and may “evaluate the persuasiveness of conflicting testimony and decide which is more likely 6 true”
Written by the judges who cited it.
The opinion
REINHARDT, Circuit Judge,
concurring in the result:
Judge Silverman’s argument that Kear-ney is entitled to summary judgment on the current record is extremely persuasive, and I cannot disagree with his analysis. However, although the question is a close one, I conclude that the litigation may benefit from a trial on the merits. Given the absence of evidence in the record regarding the duties of a trial attorney, as well as the uncertain and conflicting nature of some of the medical evidence, a genuine issue of material fact may well remain as to whether Kearney’s limitations disabled him from his “specialty” in the practice of law.
While I can concur in the decision to remand for a trial and in Section I of what Judge Kleinfeld contends is a majority opinion, I unfortunately cannot concur in the remainder of that opinion. At places, the opinion appears to me to be unclear or *1097 uncertain and, as a result, is.susceptible of conveying an inaccurate message. In other places, it contains dicta that is not only unnecessary but incorrect. In addition, I disagree with some of Judge Kleinfeld’s legal and philosophical conclusions, such as his statement that a conflict existed between Kearney and Snow 1 and that a less “holistic” approach may be the solution to our problem.
I also agree with Judge Fletcher’s characterization of Judge Kleinfeld’s opinion as a plurality opinion, at least with respect to Parts III and IV. I do not think that votes by dissenting judges can be used to make a plurality opinion a majority one. Judge O’Seannlain has joined the dissenting opinion but also seeks to concur separately in Parts III and IV. His effort does not change the status of Judge Kleinfeld’s opinion.
Rather than try to identify the precise statements in Judge Kleinfeld’s opinion with which I agree or disagree, I will try to restate what I believe our remand means. Let me begin by saying that our remand does not limit the discretion of the district court to consider evidence outside the administrative record or recommend that the district judge decline to accept evidence that supplements that record. We simply restate the rule of Mongeluzo v. Baxter Travenol Long Term Disability Benefit Plan, 46 F.3d 938 (9th Cir.1995), and hold that the district court did not abuse its discretion under that standard at the time of the summary judgment proceeding. Given that, in the summary judgment proceeding, Kearney offered in evidence only a brief definition from the Occupational Outlook Handbook and failed to specify any additional medical evidence he sought to offer beyond that already in the record before the administrator, I do not disagree with the conclusion that, for the purposes of summary judgment, the district court correctly limited the evidence to that record.
Obviously, we cannot issue rulings at this stage of the proceedings on the admissibility of evidence not yet presented to the district court, just as we cannot rule in advance on the standard of review that will be applied if the district court’s decision on the merits after trial becomes the subject of a future appeal. At trial, the district court will face a new and very different problem (even though the ultimate question, whether in view of his physical and mental limitations Kearney can properly continue to practice his specialty as a trial lawyer, remains the same). Judge Klein-feld’s opinion acknowledges that we are not ruling in advance on the admissibility of any additional evidence that the parties may seek to present at trial; that the district court will now be required to make findings of fact and answer a different question in its review of the record; and that the district court “will retain discretion to decide, subject to Mongeluzo , whether, in order to answer this different question, he should take additional evidence.” Kleinfeld Op. at 3890-91.
Of course, we cannot predict what additional evidence Kearney will submit and whether such evidence will meet Monge-luzo ’s standard. Kearney may offer evi *1098 dence regarding the duties of a trial attorney that the district court finds more useful than the brief definition contained in the Occupational Outlook Handbook. He may also offer specific medical evidence that could not have been presented to the plan administrator. Kearney may also seek to offer other additional evidence that he did not believe necessary at the time Standard was considering his claim, but that would be pertinent at time of trial. And finally, as Judge Kleinfeld’s opinion points out, the district court may decide that evidence that was previously unnecessary is now essential for the different kind of review that it will engage in on remand. We simply cannot foresee what the parties will present or argue, and we do not know how the district court will rule on such evidence when it conducts its trial on the merits.
In fact, it is quite clear that additional evidence will be required in order to resolve the question whether Kearney’s disability precludes him from performing the material duties of his specialty as a trial lawyer. Even though the district court was within its authority to reject Kear-ney’s evidentiary offer of the Handbook definition, there is a serious gap in the record regarding the duties of a trial attorney. The experience of one judge, which led him to conclude that the need to “avoid severe emotional stress or very hard work does not warrant the conclusion that [Kearney] cannot practice as a trial attorney,” 2 is clearly not sufficient. Those of us who have been trial lawyers, particularly those familiar with the obligations of plaintiffs’ trial lawyers, and the severe stress involved when one holds the future of an injured client and his family, or indeed a widow and hers, in one’s hands, know that emotional stress, time pressure, long hours, a tight time schedule, and extremely hard work are an unavoidable part of trial practice. But rather than relying on his own or any other judge’s personal experience, the district court should admit appropriate additional evidence that is offered on this question.
The factors that courts must consider in determining whether to allow additional evidence are set forth in Quesinberry, the case that Mongeluzo followed:
claims that require consideration of complex medical questions or issues regarding the credibility of medical experts; the availability of very limited administrative review processes with little or no evidentiary record; the necessity of evidence regarding interpretation of the terms of the plan rather than specific historical facts; instances where the payor and the administrator are the same entity and the court is concerned about impartiality; ... and circumstances in which there is additional evidence that the claimant could not have presented in the administrative process.
Quesinberry v. Life Ins. Co., 987 F.2d 1017 , 1026-27 (4th Cir.1993) (en banc) (emphasis added). Kearney’s claim turns on the interpretation of a key provision of the plan, the definition of the “material duties” of a trial attorney. The plan administrator’s erroneous conclusion that the nature of a trial attorney’s work was “sedentary;” the lack of notice to Kearney that Standard would adopt such an erroneous definition and rely on it in denying his claim; the materiality of this dispute to the resolution of Kearney’s claim; the lack of any information in the evidentiary record on this question; and the effect that Standard’s conflict of interest may have had on its decision-making; together necessitate the admission of additional evidence on this question at a trial on the merits. See Mongeluzo, 46 F.3d at 944 ; Quesinberry, 987 F.2d at 1026-27.
I agree with Judge Kleinfeld’s conclusion that a full trial de novo is not necessary, and that the district court should begin with the record that was before Standard and then determine what “addi *1099 tional evidence is necessary to conduct an adequate de novo review.” Mongeluzo, 46 F.3d at 944 . See also Quesinberry, 987 F.2d at 1025. But indeed it would make no sense to remand simply to apply a different and more difficult standard— from the appellant’s standpoint — to the identical record. 3 If I believed that the record before the district court was adequate, and that the district court should simply review that record without taking additional evidence, then I would agree with others on both sides that there is absolutely no purpose to a remand. I would then join Judge Silverman’s persuasive opinion.
Judge Kleinfeld also includes a brief prediction that, after the district court decides this case on remand, we will review the district court’s findings of fact under a clearly erroneous standard, regardless of whether the trial judge admits additional evidence, hears testimony, or merely confines his review to the record that was before the plan administrator. However, we, of course, cannot determine in advance what the standard of review will be for a case that is not before us. If the district court’s findings and conclusions are based solely on the record that was before the plan administrator, there is a strong argument that our review need not be so deferential. “District court review of agency action is generally accorded no particular deference, because the district court, limited to the administrative record, is in no better position to review the agency than the court of appeals.” Great Western Bank v. Office of Thrift Supervision, 916 F.2d 1421, 1426 (9th Cir.1990) (quoting Asarco, Inc. v. EPA, 616 F.2d 1153, 1161 (9th Cir.1980)). Similarly, “[bjecause we are in as good a position as the district court to review the findings of the bankruptcy court, we independently review the bankruptcy court’s decision.” In re Marquam Inv. Corp., 942 F.2d 1462, 1465 (9th Cir.1991) (quoting Matter of Pizza of Hawaii, Inc., 761 F.2d 1374, 1377 (9th Cir.1985)). Likewise in social security disability cases we independently review the administrative law judge’s decision. See Sousa v. Callahan, 143 F.3d 1240, 1243 (9th Cir.1998). Were the district court in this instance to limit its review to the record — which it should not — the question whether to afford deference to its factual findings would be an open one. This, of course, is wholly aside from the question whether the refusal to admit the evidence would be reversible as an abuse of discretion. Because at this stage of the proceedings, we do not even know what it is that we will be reviewing, if anything, any discussion as to the applicable standard of review is, at most, dictum.
. In fact, there was no such conflict. Conflicts exist between opinions, not between the arguments and briefs of the parties. There is no conflict between the two opinions because while the Kearney panel opinion, like Judge Kleinfeld’s en banc opinion, carefully sets forth and construes the specific language of a policy provision, no policy provision is set forth in Snow, and the Snow opinion therefore contains no construction of any identifiable policy language, regardless of what the parties' briefs represent about the underlying facts of that case. Cf. Kearney v. Standard Ins. Co., 144 F.3d 597, 601 (9th Cir.1998), reh’g granted, opinion withdrawn by 152 F.3d 1098 (9th Cir.1998) with Snow v. Standard Ins. Co., 87 F.3d 327, 330 (9th Cir.1996). As Section I of Judge Kleinfeld’s opinion makes clear, it is the precise language of the policy that is critical when we attempt to analyze the legal issues involved in determining coverage. For that reason, contrary to Judge Kleinfeld's earlier statement, Snow was of little or no precedential value, and Kearney was indisputably not in conflict with it.
. As Judge Kleinfeld's opinion notes, Dr. Weinreb had warned Kearney to avoid "very heavy work and very severe emotional stress” or working on a "very tight lime schedule.”
. Contrary to the optimistic ruminations of Judge Kleinfeld, I believe that a judge who granted summary judgment to the defendant on the record before us would necessarily grant that party judgment on the merits, were he to restrict his review to that record but this time be free to draw all inferences in favor of the defendant rather than the plaintiff.